Chris, here is my problem with PIOMAS and let me see if I can get it exactly right.
If we take the season from 2011 maximum to 2011 minimum, to 2012 maximum, to 2012 minimum, to 2013 maximum.
Then we compare max PIOMAS
Then we go to the CT area and confirm the comparable maximum area and the comparable minimum area in the following
2011 max/min
2012 max/min
2013 max
We should be able to see the corresponding volume loss of the 2012 huge extra melt reflected in the 2013 max ice PIOMAS.
Because whatever we melted more than the 2011 minimum must have been more than first year ice. Let us be generous and say 2.5m instead of 2m FYI which will be created to replace that which was lost in that 2012 record area minimum.
Let's be even more generous and assume that the difference between 2012 max area and 2013 max area (2013 being bigger), is actually MYI instead of FYI. I'm doing this because it makes the calculations easier for me.
What we should see, were PIOMAS estimating correctly, is the difference between the MYI area lost in the exceptional melt and the FYI ice area which replaced it at the 2013 max.
Follow me so far?
So to the calculations.
2011 Max area was 12.99m mkm^2 on April 6th
2011 Min area was 2.9899m mkm^2 on 24 Aug
2012 Max area was 13.70m mkm^2 on on 28 March
2012 Min area was 2.6399m mkm^2 on on 26 Sept
2013 Max area was 13.79m mkm^2 on on 26 Feb
Although this area peaked again at 13.79 on 14 Mar
OK so to PIOMAS
2011 max volume was 22.5mkm^3 average for April
2012 max volume was 23.11mkm^3 average for April
2013 max volume was 23.12mkm^3 average for April
Now here's the thing that blows me away and makes me fail to trust PIOMAS
2011 - 2012 I can understand. It was clearly more volume and it was roughly in April that the volume was largest.
OK so why 2013
2012 had an area low 350kkm^2 lower than 2011. That means that 350kkm^2 of MYI melted to be replaced with FYI.
OK on my rules above, 2013 was 90kkm^2 higher than 2012 so we remove that and get a figure of 260kkm^2 more FYI than MYI in the 2013 maximum, compared to the 2012 maximum.
So at the postulated 0.5m thickness, that's 130kkm^3 additional volume lost in the 2012 summer melt which has to be replaced by the 2013 re-freeze.
OK, again, that is within the bounds of variance for a colder winter. Except that the winter was exceptionally slow to start for 2012/13 in terms of area. 2012 was lower than 2011 almost all the way to the year end. 2013 started much lower than 2012 until mid Jan where they crossed over.
Which means, the FYI in 2013 was thinner than the FYI in 2012.
So I don't understand how we get a max ice volume some 10kkm^3 larger than 2012. All logic tells me that it should be at least the 130kkm^3 lower if not more because of the slow FYI generation. Although I could be wrong on that.
Finally I can't align the PIOMAS data with the area calculations over the months.
2013 peaked on Feb 26th then remained high till Mar 14th. It then fell from that date all the way to the end of April. In fact April ended 1.85mkm^2 lower than the 14th March maximum.
Yet the PIOMAS figures for Feb, Mar and April 2013 were 19.318, 21.964 and 23.122 respectively.
The only answer I have for this is "It's a model".
But my point is this. If it can get that situation wrong, then what did it do with the rest of it and since. It may be that the averaging mechanism within the model does balance out.
But here's the final kicker. A lot of what melted out in 2012 was not 2.5m ice. There was a lot of 3,4 and 5 year ice vanished that year. It wasn't 350*0.5 It was a hell of a lot more. Also it was not predominantly 2m FYI ice which replaced the lost ice from 2012 but a lot of 1M and 1.5M ice at the periphery. I do recall that because of this very thin ice there was an expectation that 2014 would be another serious loss year.
If we add all this together, it should mean that PIOMAS by now is seriously off in it's calculation.
Or I got it all wrong and there is a simple explanation as to why FYI on a short warm winter can replace all the MYI volume lost in an exceptional melt year?